Nonconvex, Fully Distributed Optimization Based CAV Platooning Control Under Nonlinear Vehicle Dynamics

CAV platooning technology has received considerable attention, driven by the next generation smart transportation systems. This paper considers nonlinear vehicle dynamics and develops fully distributed optimization based control schemes via platoon centered MPC approach for a possibly heterogeneous platoon. The leads to major difficulties in algorithm development analysis. Specifically, underlying problem is nonconvex densely coupled. Further, closed loop becomes time-varying system with non-vanishing external perturbations, making stability analysis rather complicated. To overcome these difficulties, we formulate as locally coupled, albeit nonconvex, develop sequential convex programming scheme general horizon. Such can be effectively implemented real-time computing using operator splitting methods. analyze stability, apply various tools from global implicit function theorems, of linear systems, Lyapunov theory input-to-state show that stable uniformly all small coefficients pertaining dynamic effects. Numerical tests on real traffic condition illustrate effectiveness proposed method.